> load('.deploy_info.Rdata')
> rsconnect::setAccountInfo(name=deploy_info$name, token=deploy_info$token,
+ secret=deploy_info$secret)
> rsconnect::deployApp(appFiles = c('ui.R', 'server.R', 'meth_df_withSymbol.Rdata',
+ 'meth_data.Rdata', 'global.R',
+ 'tf_data.Rdata', 'google-analytics.js', 'www/LICENSE.txt'),
+ appName = 'wgbsExprs', account = 'jhubiostatistics', server = 'shinyapps.io')
Preparing to deploy application...Update application currently deployed at
https://jhubiostatistics.shinyapps.io/wgbsExprs/? [Y/n] Y
DONE
Uploading bundle for application: 302754...--- Please select a CRAN mirror for use in this session ---
DONE
Deploying bundle: 2448146 for application: 302754 ...
Waiting for task: 652014864
building: Parsing manifest
################################ Begin Task Log ################################
################################# End Task Log #################################
Error: Unhandled Exception: Child Task 652014866 failed: Error parsing manifest: Unable to determine package source for Bioconductor package GenomicAlignments: Repository must be specified
> traceback()
3: stop(status$error, call. = FALSE)
2: client$waitForTask(taskId, quiet)
1: rsconnect::deployApp(appFiles = c("ui.R", "server.R", "meth_df_withSymbol.Rdata",
"meth_data.Rdata", "global.R", "tf_data.Rdata", "google-analytics.js",
"www/LICENSE.txt"), appName = "wgbsExprs", account = "jhubiostatistics",
server = "shinyapps.io")
From the error message, https://github.com/rstudio/rsconnect/issues/372 and https://github.com/rstudio/rsconnect/issues/88 might be related. In particular, due to https://github.com/rstudio/rsconnect/issues/88 I re-installed GenomicAlignments using BiocManager::install('GenomicAlignments'). Though this makes me think that rsconnect might have issues with Bioconductor packages in general and the time frame is consistent with the change in how Bioconductor packages are installed (biocLite('pkg') was the norm for years, vs BiocManager::install('pkg') used now) and might be related to https://github.com/rstudio/rsconnect/issues/274. Reading https://github.com/rstudio/rsconnect/issues/372 prompted me to comment out options(repos = c(CRAN = "https://cran.rstudio.com")) in my ~/.Rprofile file, but it had no effect. I also tried using the latest rsconnect from GitHub but that failed too (on the repositories prompt, I select the cloud one at the top of the list). Furthermore, using RSwitch I went back to my R 3.4.4 installation (not 3.4.3 like the version that is currently deployed; I don't have that anymore) and tried from there, but I first had to update my rsconnect given the following error:
> rsconnect::setAccountInfo(name=deploy_info$name, token=deploy_info$token,
+ secret=deploy_info$secret)
Error: HTTP 426
GET https://api.shinyapps.io/v1/users/current/
The rsconnect package is out of date.
Please update to the latest available on CRAN.
and then ran into the same initial error I see on R 3.6.1:
> rsconnect::deployApp(appFiles = c('ui.R', 'server.R', 'meth_df_withSymbol.Rdata',
+ 'meth_data.Rdata', 'global.R',
+ 'tf_data.Rdata', 'google-analytics.js', 'www/LICENSE.txt'),
+ appName = 'wgbsExprs', account = 'jhubiostatistics', server = 'shinyapps.io')
Preparing to deploy application...Update application currently deployed at
https://jhubiostatistics.shinyapps.io/wgbsExprs/? [Y/n] Y
DONE
Uploading bundle for application: 302754...DONE
Deploying bundle: 2448133 for application: 302754 ...
Waiting for task: 652013408
building: Processing bundle: 2448133
building: Parsing manifest
################################ Begin Task Log ################################
################################# End Task Log #################################
Error: Unhandled Exception: Child Task 652013409 failed: Error parsing manifest: Unable to determine package source for Bioconductor package GenomicAlignments: Repository must be specified
Sooo.... I'm not sure how to proceed here. This is using a Pro ShinyApps account from the JHU Biostatistics Department if that matters at all. Though I doubt it does based on the other issues I linked to.
If it's relevant, note that using shiny::runApp() works well in R 3.6.1.
If you need any other information, please let me know.
Hi,
Months ago (over a year ago) I deployed https://jhubiostatistics.shinyapps.io/wgbsExprs/ from files found here https://github.com/LieberInstitute/brain-epigenomics/tree/master/meth_vs_expr/meth_shiny with R 3.4.3. I'm now trying to update the deployed app using R 3.6.1 with some simple changes to the
ui.R
file (the citation information for the shiny app since it's tied to a paper that just got published π) but it's breaking πwith:Clean code at
deploy.R
From the error message, https://github.com/rstudio/rsconnect/issues/372 and https://github.com/rstudio/rsconnect/issues/88 might be related. In particular, due to https://github.com/rstudio/rsconnect/issues/88 I re-installed
GenomicAlignments
usingBiocManager::install('GenomicAlignments')
. Though this makes me think thatrsconnect
might have issues with Bioconductor packages in general and the time frame is consistent with the change in how Bioconductor packages are installed (biocLite('pkg')
was the norm for years, vsBiocManager::install('pkg')
used now) and might be related to https://github.com/rstudio/rsconnect/issues/274. Reading https://github.com/rstudio/rsconnect/issues/372 prompted me to comment outoptions(repos = c(CRAN = "https://cran.rstudio.com"))
in my~/.Rprofile
file, but it had no effect. I also tried using the latestrsconnect
from GitHub but that failed too (on the repositories prompt, I select the cloud one at the top of the list). Furthermore, usingRSwitch
I went back to my R 3.4.4 installation (not 3.4.3 like the version that is currently deployed; I don't have that anymore) and tried from there, but I first had to update myrsconnect
given the following error:and then ran into the same initial error I see on R 3.6.1:
Sooo.... I'm not sure how to proceed here. This is using a Pro ShinyApps account from the JHU Biostatistics Department if that matters at all. Though I doubt it does based on the other issues I linked to.
If it's relevant, note that using
shiny::runApp()
works well in R 3.6.1.If you need any other information, please let me know.
Best, Leo
R session info
Currently deployed (2019-09-26)
The documentation tab of https://jhubiostatistics.shinyapps.io/wgbsExprs/ prints the sesion information, which I see is:
I don't have the version number for
rsconnect
saved from the time that I deployed the app.Current deploy attempt
This is the R session info after trying to deploy with
rsconnect::deployApp()
and then confirming that the app works withshiny::runApp()
.Here's the R session information from my last deploy attempt using
rsconnect
's GitHub version instead of the CRAN one.Finally, here's the R session information from my attempt with R 3.4.4 and a new
rsconnect
: